Designing resilient applications for Azure

In a distributed system, failures will happen. Hardware can fail. The network can have transient failures. Rarely, an entire service or region may experience a disruption, but even those must be planned for.

Building a reliable application in the cloud is different than building a reliable application in an enterprise setting. While historically you may have purchased higher-end hardware to scale up, in a cloud environment you must scale out instead of scaling up. Costs for cloud environments are kept low through the use of commodity hardware. Instead of trying to prevent failures altogether, the goal is to minimize the effects of a failure within the system.

This article provides an overview of how to build resilient applications in Microsoft Azure. It starts with a definition of the term resiliency and related concepts. Then it describes a process for achieving resiliency, using a structured approach over the lifetime of an application, from design and implementation to deployment and operations.

What is resiliency?

Resiliency is the ability of a system to recover from failures and continue to function. It's not about avoiding failures, but responding to failures in a way that avoids downtime or data loss. The goal of resiliency is to return the application to a fully functioning state following a failure.

Two important aspects of resiliency are high availability and disaster recovery.

  • High availability (HA) is the ability of the application to continue running in a healthy state, without significant downtime. By "healthy state," we mean the application is responsive, and users can connect to the application and interact with it.
  • Disaster recovery (DR) is the ability to recover from rare but major incidents: non-transient, wide-scale failures, such as service disruption that affects an entire region. Disaster recovery includes data backup and archiving, and may include manual intervention, such as restoring a database from backup.

One way to think about HA versus DR is that DR starts when the impact of a fault exceeds the ability of the HA design to handle it.

When you design resiliency, you must understand your availability requirements. How much downtime is acceptable? This is partly a function of cost. How much will potential downtime cost your business? How much should you invest in making the application highly available? You also have to define what it means for the application to be available. For example, is the application "down" if a customer can submit an order but the system cannot process it within the normal timeframe? Also consider the probability of a particular type of outage occurring, and whether a mitigation strategy is cost-effective.

Another common term is business continuity (BC), which is the ability to perform essential business functions during and after adverse conditions, such as a natural disaster or a downed service. BC covers the entire operation of the business, including physical facilities, people, communications, transportation, and IT. This article focuses on cloud applications, but resilience planning must be done in the context of overall BC requirements.

Data backup is a critical part of DR. If the stateless components of an application fail, you can always redeploy them. But if data is lost, the system can't return to a stable state. Data must be backed up, ideally in a different region in case of a region-wide disaster.

Backup is distinct from data replication. Data replication involves copying data in near-real-time, so that the system can fail over quickly to a replica. Many databases systems support replication; for example, SQL Server supports SQL Server Always On Availability Groups. Data replication can reduce how long it takes to recover from an outage, by ensuring that a replica of the data is always standing by. However, data replication won't protect against human error. If data gets corrupted because of human error, the corrupted data just gets copied to the replicas. Therefore, you still need to include long-term backup in your DR strategy.

Process to achieve resiliency

Resiliency is not an add-on. It must be designed into the system and put into operational practice. Here is a general model to follow:

  1. Define your availability requirements, based on business needs.
  2. Design the application for resiliency. Start with an architecture that follows proven practices, and then identify the possible failure points in that architecture.
  3. Implement strategies to detect and recover from failures.
  4. Test the implementation by simulating faults and triggering forced failovers.
  5. Deploy the application into production using a reliable, repeatable process.
  6. Monitor the application to detect failures. By monitoring the system, you can gauge the health of the application and respond to incidents if necessary.
  7. Respond if there are failure that require manual interventions.

In the remainder of this article, we discuss each of these steps in more detail.

Define your availability requirements

Resiliency planning starts with business requirements. Here are some approaches for thinking about resiliency in those terms.

Decompose by workload

Many cloud solutions consist of multiple application workloads. The term "workload" in this context means a discrete capability or computing task, which can be logically separated from other tasks, in terms of business logic and data storage requirements. For example, an e-commerce app might include the following workloads:

  • Browse and search a product catalog.
  • Create and track orders.
  • View recommendations.

These workloads might have different requirements for availability, scalability, data consistency, and disaster recovery. There are business decisions to be made in terms of balancing cost versus risk.

Also consider usage patterns. Are there certain critical periods when the system must be available? For example, a tax-filing service can't go down right before the filing deadline, a video streaming service must stay up during a big sports event, and so on. During the critical periods, you might have redundant deployments across several regions, so the application could fail over if one region failed. However, a multi-region deployment is potentially more expensive, so during less critical times, you might run the application in a single region. In some cases, the additional expense can be mitigated by using modern serverless techniques, which use consumption-based billing, so you are not charged for under-utilitzed compute resources.


Two important metrics to consider are the recovery time objective and recovery point objective, as they pertain to disaster recovery.

  • Recovery time objective (RTO) is the maximum acceptable time that an application can be unavailable after an incident. If your RTO is 90 minutes, you must be able to restore the application to a running state within 90 minutes from the start of a disaster. If you have a very low RTO, you might keep a second regional deployment continually running an active/passive configuration on standby, to protect against a regional outage. In some cases you might deploy an active/active configuration to achieve even lower RTO.

  • Recovery point objective (RPO) is the maximum duration of data loss that is acceptable during a disaster. For example, if you store data in a single database, with no replication to other databases, and perform hourly backups, you could lose up to an hour of data.

RTO and RPO are non-functional requirements of a system, and should be dictated by business requirements. To derive these values, it's a good idea to conduct a risk assessment, and clearly understanding the cost of downtime or data loss.


Two other common measures of availability are mean time to recover (MTTR) and mean time between failures (MTBF). These measures are usually used internally by service providers to determine where to add redundancy to cloud services, and which SLAs to provide to customers.

Mean time to recover (MTTR) is the average time that it takes to restore a component after a failure. MTTR is an empirical fact about a component. Based on the MTTR of each component, you can estimate the MTTR of an entire application. Building applications from multiple components with low MTTR values results in an application with a low overall MTTR — one that recovers quickly from failures.

Mean time between failures (MTBF) is the runtime that a component can reasonably expect to last between outages. This metric can help you to calculate how frequently a service will become unavailable. An unreliable component has a low MTBF, resulting in a low SLA number for that component. However, a low MTBF can be mitigated by deploying multiple instances of the component and implementing failover between them.


If ANY of the MTTR values of components in a high-availability setup exceed the RTO of the system, then a failure in the system will cause an unacceptable business disruption. It won't be possible to restore the system within the defined RTO.


In Azure, the Service Level Agreement (SLA) describes Microsoft’s commitments for uptime and connectivity. If the SLA for a particular service is 99.9%, it means you should expect the service to be available 99.9% of the time.


The Azure SLA also includes provisions for obtaining a service credit if the SLA is not met, along with specific definitions of "availability" for each service. That aspect of the SLA acts as an enforcement policy.

You should define your own target SLAs for each workload in your solution. An SLA makes it possible to evaluate whether the architecture meets the business requirements. Perform a dependency mapping exercise to identify internal and external dependencies, such as Active Directory or third-party services such as a payment provider or e-mail messaging service. In particular, pay attention to any external dependencies that can be single point of failure or cause bottlenecks during an event. For example, if a workload requires 99.99% uptime, but depends on a service with a 99.9% SLA, that service cannot be a single-point of failure in the system. One remedy is to have a fallback path in case the service fails, or take other measures to recover from a failure in that service.

The following table shows the potential cumulative downtime for various SLA levels.

SLA Downtime per week Downtime per month Downtime per year
99% 1.68 hours 7.2 hours 3.65 days
99.9% 10.1 minutes 43.2 minutes 8.76 hours
99.95% 5 minutes 21.6 minutes 4.38 hours
99.99% 1.01 minutes 4.32 minutes 52.56 minutes
99.999% 6 seconds 25.9 seconds 5.26 minutes

Of course, higher availability is better, everything else being equal. But as you strive for more 9s, the cost and complexity to achieve that level of availability grows. An uptime of 99.99% translates to about 5 minutes of total downtime per month. Is it worth the additional complexity and cost to reach five 9s? The answer depends on the business requirements.

Here are some other considerations when defining an SLA:

  • To achieve four 9's (99.99%), you probably can't rely on manual intervention to recover from failures. The application must be self-diagnosing and self-healing.
  • Beyond four 9's, it is challenging to detect outages quickly enough to meet the SLA.
  • Think about the time window that your SLA is measured against. The smaller the window, the tighter the tolerances. It probably doesn't make sense to define your SLA in terms of hourly or daily uptime.
  • Consider the MTBF and MTTR measurements. The lower your SLA, the less frequently the service can go down, and the quicker the service must recover.

Composite SLAs

Consider an App Service web app that writes to Azure SQL Database. At the time of this writing, these Azure services have the following SLAs:

  • App Service Web Apps = 99.95%
  • SQL Database = 99.99%

Composite SLA

What is the maximum downtime you would expect for this application? If either service fails, the whole application fails. In general, the probability of each service failing is independent, so the composite SLA for this application is 99.95% × 99.99% = 99.94%. That's lower than the individual SLAs, which isn't surprising, because an application that relies on multiple services has more potential failure points.

On the other hand, you can improve the composite SLA by creating independent fallback paths. For example, if SQL Database is unavailable, put transactions into a queue, to be processed later.

Composite SLA

With this design, the application is still available even if it can't connect to the database. However, it fails if the database and the queue both fail at the same time. The expected percentage of time for a simultaneous failure is 0.0001 × 0.001, so the composite SLA for this combined path is:

  • Database OR queue = 1.0 − (0.0001 × 0.001) = 99.99999%

The total composite SLA is:

  • Web app AND (database OR queue) = 99.95% × 99.99999% = ~99.95%

But there are tradeoffs to this approach. The application logic is more complex, you are paying for the queue, and there may be data consistency issues to consider.

SLA for multi-region deployments. Another HA technique is to deploy the application in more than one region, and use Azure Traffic Manager to fail over if the application fails in one region. For a multi-region deployment, the composite SLA is calculated as follows.

Let N be the composite SLA for the application deployed in one region, and R be the number of regions where the application is deployed. The expected chance that the application will fail in all regions at the same time is ((1 − N) ^ R).

For example, if the single-region SLA is 99.95%,

  • The combined SLA for two regions = (1 − (1 − 0.9995) ^ 2) = 99.999975%
  • The combined SLA for four regions = (1 − (1 − 0.9995) ^ 4) = 99.999999%

You must also factor in the SLA for Traffic Manager. At the time of this writing, the SLA for Traffic Manager SLA is 99.99%.

Also, failing over is not instantaneous in active-passive configurations, which can result in some downtime during a failover. See Traffic Manager endpoint monitoring and failover.

The calculated SLA number is a useful baseline, but it doesn't tell the whole story about availability. Often, an application can degrade gracefully when a non-critical path fails. Consider an application that shows a catalog of books. If the application can't retrieve the thumbnail image for the cover, it might show a placeholder image. In that case, failing to get the image does not reduce the application's uptime, although it affects the user experience.

Design for resiliency

During the design phase, you should perform a failure mode analysis (FMA). The goal of an FMA is to identify possible points of failure, and define how the application will respond to those failures.

  • How will the application detect this type of failure?
  • How will the application respond to this type of failure?
  • How will you log and monitor this type of failure?

For more information about the FMA process, with specific recommendations for Azure, see Azure resiliency guidance: Failure mode analysis.

Example of identifying failure modes and detection strategy

Failure point: Call to an external web service / API.

Failure mode Detection strategy
Service is unavailable HTTP 5xx
Throttling HTTP 429 (Too Many Requests)
Authentication HTTP 401 (Unauthorized)
Slow response Request times out

Redundancy and designing for failure

Failures can vary in the scope of their impact. Some hardware failures, such as a failed disk, may affect a single host machine. A failed network switch could affect a whole server rack. Less common are failures that disrupt a whole datacenter, such as loss of power in a datacenter. Rarely, an entire region could become unavailable.

One of the main ways to make an application resilient is through redundancy. But you need to plan for this redundancy when you design the application. Also, the level of redundancy that you need depends on your business requirements — not every application needs redundancy across regions to guard against a regional outage. In general, there is a tradeoff between greater redundancy and reliability versus higher cost and complexity.

Azure has a number of features to make an application redundant at every level of failure, from an individual VM to an entire region.

Azure resiliency features

Single VM. Azure provides an uptime SLA for single VMs. (The VM must use premium storage for all Operating System Disks and Data Disks.) Although you can get a higher SLA by running two or more VMs, a single VM may be reliable enough for some workloads. For production workloads, however, we recommend using two or more VMs for redundancy.

Availability sets. To protect against localized hardware failures, such as a disk or network switch failing, deploy two or more VMs in an availability set. An availability set consists of two or more fault domains that share a common power source and network switch. VMs in an availability set are distributed across the fault domains, so if a hardware failure affects one fault domain, network traffic can still be routed the VMs in the other fault domains. For more information about Availability Sets, see Manage the availability of Windows virtual machines in Azure.

Availability zones. An Availability Zone is a physically separate zone within an Azure region. Each Availability Zone has a distinct power source, network, and cooling. Deploying VMs across availability zones helps to protect an application against datacenter-wide failures. Not all regions support Availability Zones. For a list of supported regions and services, see What are Availability Zones in Azure?.

If you are planning to use Availability Zones in your deployment, first validate that your application architecture and code base can support this configuration. If you are deploying commercial off-the-shelf software, consult with the software vendor and test adequately before deploying into production. An application must be able to maintain state and prevent loss of data during an outage within the configured zone. The application must support running in an elastic and distributed infrastructure with no hard-coded infrastructure components specified in the code base.

Azure Site Recovery. Replicate Azure virtual machines to another Azure region for business continuity and disaster recovery needs. You can conduct periodic DR drills to ensure you meet the compliance needs. The VM will be replicated with the specified settings to the selected region so that you can recover your applications in the event of outages in the source region. For more information, see Replicate Azure VMs using ASR. Consider the RTO and RPO numbers for your solution here and ensure that when testing, the recovery time and recovery point is appropriate for your needs.

Paired regions. To protect an application against a regional outage, you can deploy the application across multiple regions, using Azure Traffic Manager to distribute internet traffic to the different regions. Each Azure region is paired with another region. Together, these form a regional pair. With the exception of Brazil South, regional pairs are located within the same geography in order to meet data residency requirements for tax and law enforcement jurisdiction purposes.

When you design a multi-region application, take into account that network latency across regions is higher than within a region. For example, if you are replicating a database to enable failover, use synchronous data replication within a region, but asynchronous data replication across regions.

When you select paired regions, ensure both regions have required Azure services. For a list of services by region, see Products available by region. It's also critical to select the right deployment topology for disaster recovery, especially if your RPO/RTO are short. To ensure the failover region has enough capacity to support your workload, select either an active/passive (full replica) topology or an active/active topology. Keep in mind these deployment topologies might increase complexity and cost as resources in the secondary region are pre-provisioned and may sit idle. For more information, see Deployment topologies for disaster recovery

  Availability Set Availability Zone Azure Site Recovery/Paired region
Scope of failure Rack Datacenter Region
Request routing Load Balancer Cross-zone Load Balancer Traffic Manager
Network latency Very low Low Mid to high
Virtual network VNet VNet Cross-region VNet peering

Implement resiliency strategies

This section provides a survey of some common resiliency strategies. Most of these are not limited to a particular technology. The descriptions in this section summarize the general idea behind each technique, with links to further reading.

Retry transient failures. Transient failures can be caused by momentary loss of network connectivity, a dropped database connection, or a timeout when a service is busy. Often, a transient failure can be resolved simply by retrying the request. For many Azure services, the client SDK implements automatic retries, in a way that is transparent to the caller; see Retry service specific guidance.

Each retry attempt adds to the total latency. Also, too many failed requests can cause a bottleneck, as pending requests accumulate in the queue. These blocked requests might hold critical system resources such as memory, threads, database connections, and so on, which can cause cascading failures. To avoid this, increase the delay between each retry attempt, and limit the total number of failed requests.

Diagram of retry attempts

Load balance across instances. For scalability, a cloud application should be able to scale out by adding more instances. This approach also improves resiliency, because unhealthy instances can be removed from rotation. For example:

Replicate data. Replicating data is a general strategy for handling non-transient failures in a data store. Many storage technologies provide built-in replication, including Azure Storage, Azure SQL Database, Cosmos DB, and Apache Cassandra. It's important to consider both the read and write paths. Depending on the storage technology, you might have multiple writable replicas, or a single writable replica and multiple read-only replicas.

To maximize availability, replicas can be placed in multiple regions. However, this increases the latency when replicating the data. Typically, replicating across regions is done asynchronously, which implies an eventual consistency model and potential data loss if a replica fails.

You can use Azure Site Recovery to replicate Azure virtual machines from one region to another region. Site Recovery replicates data continuously to the target region. When an outage occurs at your primary site, you fail over to secondary location

Degrade gracefully. If a service fails and there is no failover path, the application may be able to degrade gracefully while still providing an acceptable user experience. For example:

  • Put a work item on a queue, to be handled later.
  • Return an estimated value.
  • Use locally cached data.
  • Show the user an error message. (This option is better than having the application stop responding to requests.)

Throttle high-volume users. Sometimes a small number of users create excessive load. That can have an impact on other users, reducing the overall availability of your application.

When a single client makes an excessive number of requests, the application might throttle the client for a certain period of time. During the throttling period, the application refuses some or all of the requests from that client (depending on the exact throttling strategy). The threshold for throttling might depend on the customer's service tier.

Throttling does not imply the client was necessarily acting maliciously, only that it exceeded its service quota. In some cases, a consumer might consistently exceed their quota or otherwise behave badly. In that case, you might go further and block the user. Typically, this is done by blocking an API key or an IP address range. For more information, see Throttling Pattern.

Use a circuit breaker. The Circuit Breaker pattern can prevent an application from repeatedly trying an operation that is likely to fail. The circuit breaker wraps calls to a service and tracks the number of recent failures. If the failure count exceeds a threshold, the circuit breaker starts returning an error code without calling the service. This gives the service time to recover.

Use load leveling to smooth out spikes in traffic. Applications may experience sudden spikes in traffic, which can overwhelm services on the backend. If a backend service cannot respond to requests quickly enough, it may cause requests to queue (back up), or cause the service to throttle the application. To avoid this, you can use a queue as a buffer. When there is a new work item, instead of calling the backend service immediately, the application queues a work item to run asynchronously. The queue acts as a buffer that smooths out peaks in the load. For more information, see Queue-Based Load Leveling Pattern.

Isolate critical resources. Failures in one subsystem can sometimes cascade, causing failures in other parts of the application. This can happen if a failure causes some resources, such as threads or sockets, not to get freed in a timely manner, leading to resource exhaustion.

To avoid this, you can partition a system into isolated groups, so that a failure in one partition does not bring down the entire system. This technique is sometimes called the Bulkhead pattern.


  • Partition a database (for example, by tenant) and assign a separate pool of web server instances for each partition.
  • Use separate thread pools to isolate calls to different services. This helps to prevent cascading failures if one of the services fails. For an example, see the Netflix Hystrix library.
  • Use containers to limit the resources available to a particular subsystem.

Diagram of the Bulkhead pattern

Apply compensating transactions. A compensating transaction is a transaction that undoes the effects of another completed transaction. In a distributed system, it can be very difficult to achieve strong transactional consistency. Compensating transactions are a way to achieve consistency by using a series of smaller, individual transactions that can be undone at each step.

For example, to book a trip, a customer might reserve a car, a hotel room, and a flight. If any of these steps fails, the entire operation fails. Instead of trying to use a single distributed transaction for the entire operation, you can define a compensating transaction for each step. For example, to undo a car reservation, you cancel the reservation. In order to complete the whole operation, a coordinator executes each step. If any step fails, the coordinator applies compensating transactions to undo any steps that were completed.

Test for resiliency

Generally, you can't test resiliency in the same way that you test application functionality (by running unit tests and so on). Instead, you must test how the end-to-end workload performs under failure conditions which only occur intermittently.

Testing is an iterative process. Test the application, measure the outcome, analyze and address any failures that result, and repeat the process.

Fault injection testing. Test the resiliency of the system during failures, either by triggering actual failures or by simulating them. Here are some common failure scenarios to test:

  • Shut down VM instances.
  • Crash processes.
  • Expire certificates.
  • Change access keys.
  • Shut down the DNS service on domain controllers.
  • Limit available system resources, such as RAM or number of threads.
  • Unmount disks.
  • Redeploy a VM.

Measure the recovery times and verify that your business requirements are met. Test combinations of failure modes as well. Make sure that failures don't cascade, and are handled in an isolated way.

This is another reason why it's important to analyze possible failure points during the design phase. The results of that analysis should be inputs into your test plan.

Load testing. Load testing is crucial for identifying failures that only happen under load, such as the backend database being overwhelmed or service throttling. Test for peak load, using production data or synthetic data that is as close to production data as possible. The goal is to see how the application behaves under real-world conditions.

Disaster recovery drills. It is not enough if you have a good disaster recovery plan in place. You need to test it periodically to ensure your recovery plan works fine when it matters. For Azure virtual machines, you can use Azure Site Recovery to replicate and perform DR drills without affecting production applications or ongoing replication.

Deploy using reliable processes

Once an application is deployed to production, updates are a possible source of errors. In the worst case, a bad update can cause downtime. To avoid this, the deployment process must be predictable and repeatable. Deployment includes provisioning Azure resources, deploying application code, and applying configuration settings. An update may involve all three, or a subset.

The crucial point is that manual deployments are prone to error. Therefore, it's recommended to have an automated, idempotent process that you can run on demand, and re-run if something fails.

Two concepts related to resilient deployment are infrastructure as code and immutable infrastructure.

  • Infrastructure as code is the practice of using code to provision and configure infrastructure. Infrastructure as code may use a declarative approach or an imperative approach (or a combination of both). Resource Manager templates are an example of a declarative approach. PowerShell scripts are an example of an imperative approach.
  • Immutable infrastructure is the principle that you shouldn’t modify infrastructure after it’s deployed to production. Otherwise, you can get into a state where ad hoc changes have been applied, so it's hard to know exactly what changed, and hard to reason about the system.

Another question is how to roll out an application update. We recommend techniques such as blue-green deployment or canary releases, which push updates in highly controlled way to minimize possible impacts from a bad deployment.

  • Blue-green deployment is a technique where an update is deployed into a production environment separate from the live application. After you validate the deployment, switch the traffic routing to the updated version. For example, Azure App Service Web Apps enables this with staging slots.
  • Canary releases are similar to blue-green deployments. Instead of switching all traffic to the updated version, you roll out the update to a small percentage of users, by routing a portion of the traffic to the new deployment. If there is a problem, back off and revert to the old deployment. Otherwise, route more of the traffic to the new version, until it gets 100% of the traffic.

Whatever approach you take, make sure that you can roll back to the last-known-good deployment, in case the new version is not functioning. Also have a strategy in place to roll back database changes and any other changes to dependent services. If errors occur, the application logs must indicate which version caused the error.

Monitor to detect failures

Monitoring is crucial for resiliency. If something fails, you need to know that it failed, and you need insights into the cause of the failure.

Monitoring a large-scale distributed system poses a significant challenge. Think about an application that runs on a few dozen VMs — it's not practical to log into each VM, one at a time, and look through log files, trying to troubleshoot a problem. Moreover, the number of VM instances is probably not static VMs get added and removed as the application scales in and out, and occasionally an instance may fail and need to be reprovisioned. In addition, a typical cloud application might use multiple data stores (Azure storage, SQL Database, Cosmos DB, Redis cache), and a single user action may span multiple subsystems.

You can think of the monitoring process as a pipeline with several distinct stages:

Composite SLA

  • Instrumentation. The raw data for monitoring comes from a variety of sources, including application logs, operating systems performance metrics, Azure resources, Azure subscriptions and Azure tenants. Most Azure services expose metrics that you can configure to analyze and determine the cause of problems.
  • Collection and storage. Raw instrumentation data can be held in various locations and with various formats (e.g., application trace logs, IIS logs, performance counters). These disparate sources are collected, consolidated, and put into reliable data stores such as Application Insights, Azure Monitor metrics, Service Health, storage accounts and Log Analytics.
  • Analysis and diagnosis. After the data is consolidated in these different data stores, it can be analyzed to troubleshoot issues and provide an overall view of application health. Generally, you can search for the data in Application Insights and Log Analytics using Kusto queries. Azure Advisor provides recommendations with a focus on resiliency and optimization.
  • Visualization and alerts. In this stage, telemetry data is presented in such a way that an operator can quickly notice problems or trends. Examples include dashboards or email alerts. With Azure dashboards, you can build a single-pane of glass view of monitoring graphs originating from Application Insights, Log Analytics, Azure Monitor metrics and service health. With Azure Monitor alerts, you can create alerts on service health and resource health.

Monitoring is not the same as failure detection. For example, your application might detect a transient error and retry, resulting in no downtime. But it should also log the retry operation, so that you can monitor the error rate, in order to get an overall picture of application health.

Application logs are an important source of diagnostics data. Best practices for application logging include:

  • Log in production. Otherwise, you lose insight where you need it most.
  • Log events at service boundaries. Include a correlation ID that flows across service boundaries. If a transaction flows through multiple services and one of them fails, the correlation ID will help you pinpoint why the transaction failed.
  • Use semantic logging, also known as structured logging. Unstructured logs make it hard to automate the consumption and analysis of the log data, which is needed at cloud scale.
  • Use asynchronous logging. Otherwise, the logging system itself can cause the application to fail by causing requests to back up, as they block while waiting to write a logging event.
  • Application logging is not the same as auditing. Auditing may be done for compliance or regulatory reasons. As such, audit records must be complete, and it's not acceptable to drop any while processing transactions. If an application requires auditing, this should be kept separate from diagnostics logging.

For more information about monitoring and diagnostics, see Monitoring and diagnostics guidance.

Respond to failures

Previous sections have focused on automated recovery strategies, which are critical for high availability. However, sometimes manual intervention is needed.

  • Alerts. Monitor your application for warning signs that may require proactive intervention. For example, if you see that SQL Database or Cosmos DB consistently throttles your application, you might need to increase your database capacity or optimize your queries. In this example, even though the application might handle the throttling errors transparently, your telemetry should still raise an alert so that you can follow up. It is recommended to configure alerts on Azure resources metrics and diagnostics logs against the services limits and quotas thresholds. We recommend to setup alerts on metrics as they are lower latency vs. diagnostics logs. In addition, Azure is able to provide with some out-of-the-box health status through resource health which can help diagnose throttling of Azure services.
  • Failover. Configure a disaster recovery strategy for your application. The appropriate strategy will depend on your SLAs. For most scenarous, an active-passive implementation is sufficient. For more information, see Deployment topologies for disaster recovery . Most Azure services allow for either manual or automated failover. For example, in an IaaS application, use Azure Site Recovery for the web and logic tiers and SQL AlwaysOn Availability Groups for the database tier. Traffic Manager provides automated failover across regions.
  • Operational readiness testing. Perform an operational readiness test for both failover to the secondary region and failback to the primary region. Many Azure services support manual failover or test failover for disaster recovery drills. Alternatively, you can simulate an outage by shutting down or removing services.
  • Data consistency check. If a failure happens in a data store, there may be data inconsistencies when the store becomes available again, especially if the data was replicated. For Azure services that provide cross-regional replication, look at the RTO and RPO to understand the expected data loss in a failure. Review the SLAs for Azure services to understand whether cross-regional failover can be initiated manually or is initiated by Microsoft. For some services, Microsoft decides when to perform the failover. Microsoft may prioritize the recovery of data in the primary region, only failing over to a secondary region if data in the primary region is deemed unrecoverable. For example, Geo-redundant storage and Key Vault follow this model.
  • Restoring from backup. In some scenarios, restoring from backup is only possible within the same region. This is the case for Azure VMs Backup. Other Azure services provide geo-replicated backups, such as Redis Cache Geo-Replicas. The purpose of backups is to protect against accidental deletion or corruption of data, restoring the application to a functional version earlier in time. Therefore, while backups can serve as a disaster recovery solution in some cases, the inverse is not always true: Disaster recovery won't protect you against accidental deletion or corruption a data.

Document and test your disaster recovery plan. Evaluate the business impact of application failures. Automate the process as much as possible, and document any manual steps, such as manual failover or data restoration from backups. Regularly test your disaster recovery process to validate and improve the plan. Set up alerts for the Azure services consumed by your application.


This article discussed resiliency from a holistic perspective, emphasizing some of the unique challenges of the cloud. These include the distributed nature of cloud computing, the use of commodity hardware, and the presence of transient network faults.

Here are the major points to take away from this article:

  • Resiliency leads to higher availability, and lower mean time to recover from failures.
  • Achieving resiliency in the cloud requires a different set of techniques from traditional on-premises solutions.
  • Resiliency does not happen by accident. It must be designed and built in from the start.
  • Resiliency touches every part of the application lifecycle, from planning and coding to operations.
  • Test and monitor!